As artificial intelligence (AI) is gaining popularity, it’s also being increasingly used in the medical field to change the way patients are diagnosed, treated or monitored. A new study has found that new tools based on AI language models could help in diagnosing schizophrenia, which is characterised by distortions of reality and disturbances of thought and language.
The study, led by researchers from University College London (UCL), has developed new tools, based on AI language models, that can characterise subtle signatures in the speech of patients diagnosed with schizophrenia, as reported by Press Trust of India (PTI).
An AI language model is a natural language model which produces probabilities of a series of words based on the text dataset it was trained on. Some of the most popular ones include ChatGPT and Bard, but for this study, the researchers used Facebook AI Research's (FAIR) fastText.
The study, published in the journal Proceedings of the National Academy of Sciences, aimed to understand how the automated analysis of language could help doctors and scientists diagnose and assess psychiatric conditions, according to a press statement by UCL. Currently, diagnosis is based mostly on talking with patients and their loved ones.
In this study, 52 study participants - 26 having schizophrenia and 26 control – performed verbal fluency tasks, as reported by PTI. The researchers used an AI language model to analyse the answers and test whether the words people spontaneously recalled could be predicted by the AI model and whether this predictability was reduced among those with schizophrenia.
The findings showed that the answers given by control participants were more predictable by the AI model and this difference was largest in patients with more severe symptoms of schizophrenia, the statement explains. These were further supported by the second part of the study which involved brain scanning to measure brain activity in regions of the brain involved in learning and storing ‘cognitive maps’, which refer mental representation of stimuli from the physical environment that helps in storing information.
“Until very recently, the automatic analysis of language has been out of reach of doctors and scientists. However, with the advent of artificial intelligence (AI) language models such as ChatGPT, this situation is changing,” lead author Matthew Nour said in the statement. “This work shows the potential of applying AI language models to psychiatry – a medical field intimately related to language and meaning,” Nour adds.
As there is huge interest in using AI in medicine, if these tools are proven to be safe and robust, they could be deployed in the clinic within the next decade, the researcher said, as reported by PTI.